You already had built a point-based scoring in Marketo, or Salesforce but want to transfer it to MadKudu? It's possible!
MadKudu offers two types of modeling for the Customer Fit model: rule-based or decision tree-based. This article covers how to build or edit the rules of a rule-based model. For the decision tree-based model, please go to this article.
Note that these rules can only be built based on firmographic, demographic, or technographic data, and cannot include behavioral activity (visited pricing page, requested a demo), which is included in the Likelihood to Buy model.
To create a rule-based model, create a Customer Fit model, then
- Click on Model,
- Turn On the Advanced mode
- Click on the Model Base tab
- Select the rule-based option
What are rules?
The output of the model is a score that sums up all the points associated with the rules a lead or account verifies.
These rules are configured in the tab Rules, when the Model Base tab is configured on "rule-based model"
Guiding principles
- The rules are organized in Groups, containing one or more rules.
- Within a Group, a lead or account would get the points associated with the first rule it verifies.
- Different groups are used for different traits like below.
Let's take an example. In the screenshot below,
- a Public company tagged B2BSaas would get 100 + 100 = 200 points
- a Non Profit company tagged B2C would get (-30) + (-10) = -40 points
- If a company is tagged both B2B and B2C, then the company would only get the points associated with the B2B tag (=30). If you'd like the company to get the points associated with the B2B tag and the B2C tag, create another group for tag_is_b2c
Prerequisites to add/edit/delete rules
- You have the permissions of the Architect or Admin role
- You know what a Computation is
How to add a rule?
Step 1: Add a new rule group
- Go to Data Science Studio > Rules
- Click Add new rule group
- Select
- the Trait (=Computation) from the list
- the condition
- the value(s)
- Check "Case insensitive" for the condition to not take into account lower /upper case.
For example to score the same "United States" and "united states"
Keep unchecked "Case insensitive" for the rule to apply only on a text with specific lower/upper case - Enter a number of Points (Integer, positive or negative)
- (optional) Click on Add sub-condition to add an additional condition to the rule.
- Click Save Changes at the bottom of the page.
You can add as many rules as you'd like within a group and as many groups as you'd like.
⚠️ If you refresh or close the page without clicking on Save changes, your changes will be lost
Step 3: Add a new Signal with the same rule to display the information in the Signals
Head to the Signals tab and follow these instructions to add signals.
Step 4: Check the result on a sample in the Spot check tab
Head to the Spot check tab and click on Get a new dataset
Step 5: Deploy Rule
Keep these considerations in mind when deploying overrides:
- All your Leads, Contacts, or Accounts that are usually scored by MadKudu will get rescored with the next batch scoring (see when the next Analysis then Sync process should run in the Processes Page).
- Adding, editing, or deleting rules that increase or decrease prospect scores may trigger automated workflows (out of MadKudu's control) in your CRM which are based on the customer fit or lead grade score (like your MQL workflow).
When you are ready, go to the Deploy tab
- If the live model you are editing is flagged "Live as Standard", then go into the first sub tab "deploy as standard" and click Primary deploy
- If the live model you are editing is flagged "Live as Multi-fit" please contact support@madkudu.com, we'll help you with that.
Well done!
How to edit a rule?
Just edit the field, condition, or value and click Save changes at the bottom of the page
How to delete a group?
Just delete all the rules within the group and click Save changes at the bottom of the page
Any question? Help us improve our documentation! Write us here